Statistical/Design Consulting
Research is about obtaining objective and reliable evidence for meaningful, useful and effective decision making. Our goal is to provide evidence that is supported by reality in statistical modeling of research data.
Thesis and Dissertation Consulting
Design and statistical consulting are essential in thesis and doctoral dissertation preparation. We provide consultation from the planning of thesis/dissertation through proposal development to IRB approval,...
Program Evaluation/Patient
In the implementation of public health or social services programs, surveillance and continuous monitoring of such programs is essential, and ensures program continuity. We at AHRI have experts on process, impact and outcomes evaluation.
Discover reality in statistical modeling
of biomedical/ clinical search data
Epidemiologic designs are used in gathering public health data for risk identification, programs planning and outcomes investigation. Often these data are not collected from random probability sampling of the study participants, and yet we quantify the random errors in these studies. Such practice should be discouraged in evidence discovery. Therefore when random variables, which are based on probability sampling of study participants are used, only then can we quantify the random error with a probability value ( p value). Without such sampling,(probability sampling method) with exception being large registry databases studies, results of epidemiologic studies should remain at a descriptive statistics level.
Population health research often involves study populations that are not obtained through probability sampling. Often we apply statistical inference on such studies, which is inappropriate. Research is conducted on sample, and the application of p value in this circumstance serves to indicate whether or not the sample is representative of the targeted population, in order to warrant inference. At AHRI, we encourage the application of inferential statistics to sample that are obtained from probability sampling, as well as large disease registry datasets. Therefore when data are not based on probability sampling and are not derived from large registries, it is meaningless to quantify random error, but to report such findings as descriptive statistic.
The p value is not the measure of evidence but merely reflects the size of the study, which is of no clinical or public health relevance. Often negative findings reflect the small size of the study and not evidence from the data. In this context, unless the power of the study was estimated, it is inappropriate to report such a negative finding as statistically insignificant without power statement, since absence of evidence does not imply the evidence of absence.
Racial/ethnic variability in incidence, prevalence of diseases in the US as well as the outcome have been well documented. Often, this observation reflects ethnic/racial geographic clustering. At the AHRI one of our major research goals is the geographic mapping of disease. Spatial epidemiology therefore remains a significant approach to understanding disease clustering by race/ethnicity
Health issues should be addressed by data-driven policies. At AHRI, we have experts on the role of epidemiology in health policy formulation and implementation. Therefore by applying sound epidemiologic principles and methodology, as well as reality in statistical modeling, in understanding the magnitude of health and health related events, AHRI is able to inform policy development and implementation. The lack of effective health policy continues to affect our health status as perceived in racial/ethnic disparities in infant mortality. Consequently without an effective health policy to address social issues associated with prenatal care such disparities will remain un-narrowed.
